Skills/Glama/protocolprobe-mcp

protocolprobe-mcp@r1

CAUTION· install only

Publisher claims: "ead-only MCP server over ProtocolProbe's public API: dated public-record facts about servers in the official MCP registry (record checks, servers by topic, registry size and daily changes). Needs no key and calls no other server's tools." Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry GlamaPublisher protocolprobe · publishing for 11 daysInstalls 0Last detonated 5d ago
What would fix thisWatchDispute verdict

Coverage

Nothing executable in the bundle ran; the trace covers the install step only.
Tested
every file read, file written and host contacted by what ran
Not tested
use of the skill by an agent

What this means for the label: Caution here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.

Before you give it your token

It will ask for
MIN_SECRET_LENGTHConfiguration, not credentials: ENGINE_CAPTURE_TOOLS, ENGINE_FULL_TEXT, PROTOCOLPROBE_API, PROTOCOLPROBE_SELFTEST_STATE_DIR, SystemRoot, XDG_STATE_HOME
Where it talks to
protocolprobe.com (its own)Also smithery.ai, evil.example, h.example, ok.example, zapier.com, x.example, supabase.co, kp3xq9mztr7wlbyv2.io, example.org, other.example, inference.sh, your.server, schemastore.org — not the publisher's domain; check what for before connecting.
What the honeypot saw
no credential files read · no network
What we could not test
use of the skill by an agent

Connect it with a scoped credential, under the guard, and read the third-party hosts first. The code names smithery.ai, evil.example, h.example, ok.example, zapier.com, x.example, supabase.co, kp3xq9mztr7wlbyv2.io, example.org, other.example, inference.sh, your.server, schemastore.org as well as its own vendor. The guard injects MIN_SECRET_LENGTH only into requests to protocolprobe.com (and its subdomains) and refuses the rest.

$vet-skills run --secrets MIN_SECRET_LENGTH=protocolprobe.com -- <the server's command>

Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.

Is protocolprobe-mcp@r1 safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (1 scenario; nothing executable in the bundle ran; the trace covers the install step only) on 2026-11-09 and scored it 76/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Hidden zero-width characters detected; Hidden zero-width characters detected; Access to sensitive file paths. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 0 s · 1 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
0.0sprocessscript backend runs setup only
0.0sprocessscript backend runs setup only
0.0sprocessno install scripts to re-run

Static findings

MEDIUM
Access to sensitive file pathsdiscover/discover.test.mjs:150 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesassert.equal(safeRemote("file:///etc/passwd"), null);Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
HIGH
Hidden zero-width characters detecteddiscover/lib.mjs:254 · Hidden zero-width characters detected — may conceal contenttext = String(text).replace(/^/, "");Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST07
MEDIUM
MCP server using network transport (SSE/HTTP)discover/discover.test.mjs:37 · MCP server using network transport (SSE/HTTP) — data sent over networkcfg.mcpServers.smithery = { type: "http", url: "https://server.smithery.ai/@smithery/server/mcp" };Fix: Namespace tool names and keep tool descriptions literal; no instructions to the agent inside them.
AST04
MEDIUM
MCP server using network transport (SSE/HTTP)discover/discover.test.mjs:31 · MCP server using network transport (SSE/HTTP) — data sent over networkcfg.mcpServers["evil\u001b[2K\u001b[1A\r\n fake — remote https://ok.example/"] = { command: "node" };Fix: Namespace tool names and keep tool descriptions literal; no instructions to the agent inside them.
AST04
MEDIUM
npx invocation of non-well-known packagediscover/discover.test.mjs:145 · npx invocation of non-well-known package — may execute untrusted codeassert.equal(commandName("npx -y @scope/pkg"), null);Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packagediscover/lib.mjs:136 · npx invocation of non-well-known package — may execute untrusted code* carries arguments ("npx -y pkg --token=…", "node C:\srv\server.js --key=…") is never printed.Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
HIGH
Hidden zero-width characters detecteddiscover/lib.mjs:279 · Hidden zero-width characters detected — may conceal contentreturn res.replace(/^/, "");Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST07
MEDIUM
MCP server using network transport (SSE/HTTP)discover/discover.test.mjs:35 · MCP server using network transport (SSE/HTTP) — data sent over networkcfg.mcpServers.hexpath = { type: "http", url: "https://h.example/mcp/deadbeefcafebabe" };Fix: Namespace tool names and keep tool descriptions literal; no instructions to the agent inside them.
AST04
MEDIUM
npx invocation of non-well-known packagediscover/lib.mjs:230 · npx invocation of non-well-known package — may execute untrusted codeconst viaRemote = url ? null : remoteArgOf(def); // `npx mcp-remote <url>`: the package AND the remote it points atFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packagediscover/lib.mjs:186 · npx invocation of non-well-known package — may execute untrusted code* The URL `mcp-remote` is started with (npx mcp-remote https://… --header …). It is an argument, so it is never shownFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Bidirectional text override characters detecteddiscover/discover.test.mjs:32 · Bidirectional text override characters detected — may hide content directioncfg.mcpServers["name‮evil"] = { command: "node" }; // right-to-left overrideFix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
MEDIUM
URL with embedded credentialsdiscover/discover.test.mjs:44 · URL with embedded credentials — may be phishing or credential leakingremote1: { type: "http", url: "https://user:[email protected]/mcp?key=LEAK_QUERY#LEAK_FRAG", headers: { Authorization: "Bearer LEAK_HEADER" } },Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST07
MEDIUM
File type combined with upload actiondiscover/discover.mjs:166 · File type combined with upload action — agent may exfiltrate local data files…le.error(`${a.error}\n${USAGE}`); process.exit(2); } if (a.help) { console.log(`${USAGE}\n\nLists the MCP servers configured on this machine (Claude Desktop, Claude Code, Cursor, VS Code) and, offline, which are in the\nofficial MCP regist…Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
Bidirectional text override characters detecteddiscover/discover.test.mjs:70 · Bidirectional text override characters detected — may hide content directionassert.equal(/\u001b|\r|‮/.test(text), false, "raw control characters printed");Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
HIGH
Access to sensitive file pathsselftest/src/patterns.js:10 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files{ id: 'sensitive-path', re: /(~\/\.ssh|id_rsa|\.aws\/credentials|\/etc\/passwd|\.env\b)/i },Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
HIGH
Access to sensitive file pathsselftest/src/sarif.js:63 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files// value of an option (`--env-file .env`, `--config cfg.json`), which is a setting, not the server's own fileFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
npx invocation of non-well-known package.github/ISSUE_TEMPLATE/bug.yml:10 · npx invocation of non-well-known package — may execute untrusted codeattributes: { label: Tool version (npx <tool> --version), placeholder: "0.5.0" }Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
MCP server using network transport (SSE/HTTP)discover/discover.test.mjs:38 · MCP server using network transport (SSE/HTTP) — data sent over networkcfg.mcpServers.dated = { type: "http", url: "https://h.example/mcp/2025-06-18" };Fix: Namespace tool names and keep tool descriptions literal; no instructions to the agent inside them.
AST04
MEDIUM
npx invocation of non-well-known packagediscover/discover.test.mjs:20 · npx invocation of non-well-known package — may execute untrusted codeoneline: { command: "npx -y pkg --token=LEAK_CMDSPACE" },Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
MCP server using network transport (SSE/HTTP)discover/discover.test.mjs:34 · MCP server using network transport (SSE/HTTP) — data sent over networkcfg.mcpServers.shortpath = { type: "sse", url: "https://h.example/mcp/s/kP3xQ9mZtRwLbYvL/sse" };Fix: Namespace tool names and keep tool descriptions literal; no instructions to the agent inside them.
AST04
MEDIUM
npx invocation of non-well-known packagediscover/registry-match.test.mjs:141 · npx invocation of non-well-known package — may execute untrusted codeassert.ok(text.includes("npx mcp-remote → remote https://server.smithery.ai/@smithery/server/mcp"));Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Hidden zero-width characters detectedmcp/lib.test.mjs:521 · Hidden zero-width characters detected — may conceal contentfor (const s of [c.label, c.detail, c.help.text, c.help.linkLabel]) assert.doesNotMatch(s, /[\u0000-\u001f\u007f-\u009f​-‏‪-‮⁠-⁩]/, `hostile ${i}`);Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST07
MEDIUM
npx invocation of non-well-known packageselftest/selftest.mjs:31 · npx invocation of non-well-known package — may execute untrusted code// `npx mcp-remote https://...`) is refused, so the stdio lane cannot be pointed at someone else's server by another route.Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Access to sensitive file pathsmcp/lib.test.mjs:115 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files"https://protocolprobe.com/api/v1?x=1", "https://protocolprobe.com/api/v1#x", "ftp://protocolprobe.com", "file:///etc/passwd",Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
Access to sensitive file pathsmcp/package.test.mjs:142 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files….url))], { env: { PATH: process.env.PATH, SystemRoot: process.env.SystemRoot, ...env }, stdio: ["pipe", "pipe", "pipe"] }); } test("bin: the process answers initial…Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
npx invocation of non-well-known packageselftest/src/session.js:36 · npx invocation of non-well-known package — may execute untrusted code// our own fixture through `npx --package <our tarball>` and through a .cmd file: npx, cmd.exe and the server stayed).Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Bidirectional text override characters detectedmcp/lib.test.mjs:521 · Bidirectional text override characters detected — may hide content directionfor (const s of [c.label, c.detail, c.help.text, c.help.linkLabel]) assert.doesNotMatch(s, /[\u0000-\u001f\u007f-\u009f​-‏‪-‮⁠-⁩]/, `hostile ${i}`);Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
MEDIUM
npx invocation of non-well-known packagediscover/registry-match.test.mjs:125 · npx invocation of non-well-known package — may execute untrusted codetest("npx mcp-remote <url>: the URL argument is shortened and matched, never shown as written; other programs' arguments stay hidden", () => {Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Access to sensitive file pathsdiscover/package.test.mjs:48 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files…ld_process\b/, /\brequire\s*\(/, /\beval\s*\(/, /new Function\s*\(/, /\bprocess\.env\.(?!APPDATA\b)/]) assert.equal(bad.test(src), false, `${f} matches ${bad}`); …Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
npx invocation of non-well-known packageselftest/selftest.mjs:48 · npx invocation of non-well-known package — may execute untrusted code// reason, never a pass. 0.1.1 counted dim 4 as passed through `npx mcp-remote http://localhost:...` for a server thatFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
URL with embedded credentialsdiscover/registry-match.test.mjs:128 · URL with embedded credentials — may be phishing or credential leakingconst a = d(["-y", "mcp-remote", "https://user:[email protected]/@smithery/server/mcp?key=LEAKQ", "--header", "Authorization:Bearer LEAKH"]);Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST07
MEDIUM
Access to sensitive file pathsmcp/package.test.mjs:91 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesconst uses = [...src.matchAll(/process\.env[^\n]*/g)].map((m) => m[0]);Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
URL with embedded credentialsdiscover/registry-match.test.mjs:89 · URL with embedded credentials — may be phishing or credential leakingsm: { url: "https://user:[email protected]/@smithery/server/mcp?key=LEAKQ" },Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST07
MEDIUM
Access to sensitive file pathsmcp/package.test.mjs:94 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesassert.doesNotMatch(src, /\benv\s*\[|\bprocess\.env\s*\[|Object\.(keys|entries|values)\(\s*process\.env/, file);Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
Process kill commandselftest/src/session.js:38 · Process kill command — agent can terminate processes// whole process tree (taskkill /T /F) while cmd.exe still links it, before the SDK's own kill (which comes 2 s after theFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05

Cost to the agent

Instructions loaded
≈43 tok
1 file · 55 in bundle

Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.

Known vulnerabilities

Vulnerability advisories (OSV) cover npm and PyPI packages; this registry has no advisory database.

What would make this Verified

4 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%
  1. Unclaimed publisher≈ +8 on the totalClaim protocolprobe: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
  2. Hidden zero-width characters detected≈ +4 on the totalKeep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
  3. Access to sensitive file paths≈ +2 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
  4. MCP server using network transport (SSE/HTTP)≈ +2 on the totalNamespace tool names and keep tool descriptions literal; no instructions to the agent inside them.
  5. npx invocation of non-well-known package≈ +2 on the totalPin versions and sources; vendor anything fetched at install time.
  6. Bidirectional text override characters detected≈ +2 on the totalReplace encoded or assembled strings with plain literals and remove undeclared network calls.
Publish a new version to Glama, then queue it here. Not your skill? Claim the handle to get verdict changes by email.Claim protocolprobePropose a patch to the maintainer